V. Cannillo et al. /Composites: Part A 37 (2006 )23-30 93 8293633393 10 um. Fg.7. Sample7:885℃.30,42MPa,30% Fig. 9. Microstructure obtained with the following processing condition 750C. 30 min 28 MPa, 20 vol% celsian Therefore, the parameters selected for the processing conditions verification were the following: 28 MPa: celsian volume fraction: 20 vol% temperature:735℃C The so-obtained sample showed satisfactory values in terms of shrinkage(20.16% in diameter and 18.31% in thickness) and relative density compared to the previous ones. Thus, this validates the methodology adopted. The microstructure of the composite is illustrated in Fig. 8. Celsian reinforcement is well distributed in the glass matrix. with a reduced porosity compared to the other samples. However, pores morphology indicates that the densification was not fully completed Therefore. an additional treatment was scheduled at temperature of 750C, which is above 735C but below 785C, leaving unaltered all the other processing conditions The presence of circular pores on the heat treated sample(see Fig 9)indicated that the densification was complete. Similar values in terms of shrinkage and density were obtained 50 um. compared to the previous case. By using an image analysis software on SEM micrographs, the amount of porosity for (b) the last two treatments was assessed. The processing routes at 735 and 750C showed densities of 99. 22 and 98.85%o of he theoretical value, respectively. Therefore, these fabrica tion conditions can be considered suitable for the obtainment of high-quality glass-celsian composites 5 Conclusions The aim of this paper was the assessment of a nev experimental method for glass-celsian composites prep- aration using a design of experiments approach. In fact, a DOE-based methodology was exploited in order to determine the optimal conditions for the processing 10μm of such composites. The designed fabrication route allowed Fig8 Validation treatment(735"C, 30 min, 28 MPa, 20 vol% celsian): (a) the obtainment of fully dense materials(about 99%o heoretical value) and thus it was possible to identify anTherefore, the parameters selected for the processing conditions verification were the following: † temperature: 735 8C; † time: 30 min; † pressure: 28 MPa; † celsian volume fraction: 20 vol%. The so-obtained sample showed satisfactory values in terms of shrinkage (20.16% in diameter and 18.31% in thickness) and relative density compared to the previous ones. Thus, this validates the methodology adopted. The microstructure of the composite is illustrated in Fig. 8. Celsian reinforcement is well distributed in the glass matrix, with a reduced porosity compared to the other samples. However, pores morphology indicates that the densification was not fully completed. Therefore, an additional treatment was scheduled at a temperature of 750 8C, which is above 735 8C but below 785 8C, leaving unaltered all the other processing conditions. The presence of circular pores on the heat treated sample (see Fig. 9) indicated that the densification was complete. Similar values in terms of shrinkage and density were obtained compared to the previous case. By using an image analysis software on SEM micrographs, the amount of porosity for the last two treatments was assessed. The processing routes at 735 and 750 8C showed densities of 99.22 and 98.85% of the theoretical value, respectively. Therefore, these fabrica￾tion conditions can be considered suitable for the obtainment of high-quality glass–celsian composites. 5. Conclusions The aim of this paper was the assessment of a new experimental method for glass–celsian composites prep￾aration using a design of experiments approach. In fact, a DOE-based methodology was exploited in order to determine the optimal conditions for the processing of such composites. The designed fabrication route allowed the obtainment of fully dense materials (about 99% theoretical value) and thus it was possible to identify an Fig. 7. Sample 7: 885 8C, 300 , 42 MPa, 30%. Fig. 8. Validation treatment (735 8C, 30 min, 28 MPa, 20 vol% celsian): (a) low magnification and (b) higher magnification. Fig. 9. Microstructure obtained with the following processing condition: 750 8C, 30 min, 28 MPa, 20 vol% celsian. V. Cannillo et al. / Composites: Part A 37 (2006) 23–30 29